Schiff-base appended polymers for phosphate removal

Article abstract of DOI:10.1080/10610278.2017.1372581

The remediation of phosphate-contaminated water bodies and the effective treatment of hyperphosphatemia are two big challenges in which phosphate recognition could play a useful role. Here, we summarise briefly the state of the art in phosphate removal. Next, we present findings from synthetic and phosphate extraction studies that involve polymeric materials with pendent Schiff-base macrocycles designed to function as phosphate anion receptors.

Full text of DOI:10.1080/10610278.2017.1372581

Supramolecular Chemistry
ISSN: 1061-0278 (Print) 1029-0478 (Online) Journal homepage: http://www.tandfonline.com/loi/gsch20
Schiff-base appended polymers for phosphate
removal
Murat K. Deliomeroglu, Vincent M. Lynch & Jonathan L. Sessler
To cite this article: Murat K. Deliomeroglu, Vincent M. Lynch & Jonathan L. Sessler (2017):
Schiff-base appended polymers for phosphate removal, Supramolecular Chemistry, DOI:
10.1080/10610278.2017.1372581
To link to this article: http://dx.doi.org/10.1080/10610278.2017.1372581
View supplementary material
Published online: 05 Sep 2017.
Submit your article to this journal
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=gsch20
Download by: [UC Santa Barbara Library]
Date: 07 September 2017, At: 18:51

Supramolecular chemiStry, 2017
https://doi.org/10.1080/10610278.2017.1372581
Schiff-base appended polymers for phosphate removal
Murat K. Deliomeroglu, Vincent M. Lynch and Jonathan L. Sessler
Department of chemistry, the university of texas at austin, austin tX, uSa
ABSTRACT
ARTICLE HISTORY
received 18 July 2017
accepted 23 august 2017
The remediation of phosphate-contaminated water bodies and the effective treatment of
hyperphosphatemia are two big challenges in which phosphate recognition could play a useful role.
Here, we summarise briefly the state of the art in phosphate removal. Next, we present findings from
synthetic and phosphate extraction studies that involve polymeric materials with pendent Schiff-
base macrocycles designed to function as phosphate anion receptors.
KEYWORDS
anion binding; phosphate;
polymers; extraction
Introduction
donors are too small to bind F− (4). However, an early solid
state structural study of a sapphyrin-F− complex (Figure 1a)
revealed that the F− ion was bound within the core of this
larger pentapyrrolic porphyrin analogue (4). This Sapphyrin
has a core size of roughly 5.5 Å (effective diameter) and
therefore, possesses a core that is large enough to accom-
modate a bound F− anion.
Figure 1 shows two macrocyclic systems, sapphyrin and
cyclo[8]pyrrole, each with an anionic guest in its central
cavity. In this Figure, the sapphyrin-F− complex is shown on
the left (a), while the cyclo[8]pyrrole-H2SO4 salt is shown on
the right (b). While a size comparison of differently shaped
Some of the challenges associated with anion recogni-
tion derive in part from the larger size of anions relative
to their isoelectronic cations, the diverse geometries that
anions often display, and the higher energies of solvation
that must typically be overcome to extract them out of an
aqueous phase (1). Anions are generally larger than cati-
ons. Therefore, larger hosts are required for recognition of
anions (2). The F− anion, for instance, has an ionic radius
of 1.2–1.4 Å (depending on the convention employed)
(3). Protonated porphyrins with trans-nitrogen distances
of about 4 Å and a potential source of four hydrogen bond
CONTACT Jonathan l. Sessler
sessler@cm.utexas.edu
Supplemental data for this article can be accessed at https://doi.org/10.1080/10610278.2017.1372581.
© 2017 informa uK limited, trading as taylor & Francis Group

2
M. K. DELIOMEROGLU ET AL.
Figure 1. (colour online) (a) Single crystal structures of the sapphyrin-F− complex and (b) the cyclo[8]pyrrole-h2So4 salt. these structures
were originally reported by Sessler et al. (4),(6) and were redrawn using data from the cambridge crystallographic Date centre (ccDc) as
detailed in ccDc numbers 1184078 and 176189, respectively.
anions, such as F− versus SO42−, is necessarily challenging,
the tetrahedral sulfate anion, characterised by a sulfur (S)
oxygen (O) bond distance of about 1.5 Å (5), is larger than
the spherical fluoride anion, at lease in a qualitative sense.
Consistent with this notion is the finding that the sulfate
anion, but not the fluoride anion, is bound well by the
diprotonated octapyrrolic macrocycle (Figure 1), which has
a relatively large cavity diameter of roughly 7.5–7.7 Å (6).
Inorganic phosphate and its protonated anionic forms
fluoroborate complex stabilized by a C–H⋯F–B hydro-
gen bonds in aqueous media (11). They also demonstrated
the use of a cationic borane receptor for the extraction of
F− from water (11). The same group recently employed a
neutral bidentate Lewis acid system containing two anti-
mony(V) centers to ‘squeeze out’ F− ions from water, thus
overcoming the hydration barrier (12). However, to our
knowledge, there is no reported artificial receptor capable
−
of extracting H2PO4 out of water into an organic phase.
(i.e. PO43−, HPO4 and H2PO4 ) are, like sulfate, quintes-
sential tetrahedral anions. Their recognition is likely to
require not only larger hosts but also appropriately shaped
binding sites that favor selective recognition. Consistent
with this thinking is the fact that there are more tripodal
or cage-like synthetic receptors for phosphate recogni-
tion than planar cyclic systems. However, independent
of design, cognizance must be taken of the fact that ani-
ons of di- and tribasic acids, such as H3PO4 and H2SO4, are
susceptible to proton transfer (7). The receptor itself can
also undergo deprotonation, which can exacerbate issues
associated with speciation. This latter issue was elegantly
noted by Gale et al., who reported that increasing the acid-
ity of pyrrolic hydrogens within a set of his receptors led
This is true even though the H2PO4− has the same hydra-
tion enthalpy value as F− (9), for which extractants have
been reported (13, 14).
2−
−
Nature has solved the problem of phosphate anion
recognition in polar media. Some of the determinants
for affecting such biological anion recognition have
been inferred from crystallographic studies of the phos-
phate-binding protein (PBP) from E. coli bacteria (15). This
periplasmic PBP is involved in the highly selective uptake
and active transport of phosphate into bacteria cells and
binds its target anion by means of multiple well-oriented
hydrogen bonds. It is unlikely that this particular protein
will see application in the area of phosphate anion extrac-
tion due to its rather high cost.
Nevertheless, biological methods are used in waste-
water processing. For instance, the so-called ‘enhanced
biological phosphate removal’ (EBPR) process, involves
storage of phosphate as polyphosphates (poly-P) in the
form of biomass (16). The underlying mechanism, known
as ‘luxury uptake’, involves circulation of activated sludge
through anaerobic and aerobic phases, with the wastewa-
ter added during the anaerobic phase (17). By means of
the alternating anaerobic-aerobic cycles microorganisms
accumulate phosphate in the form of poly-P. Efficient phos-
phate removal is achieved by the physical removal of the
−
to deprotonation by basic anions such as F−, H2PO4 , and
benzoate, rather than anion binding (8).
In polar media, including water, anion recognition is
rendered further challenging both because of solvation
(e.g. ΔGh = −300 kJ/mol for NO3− and ΔGh = −465 kJ/mol
−
for H2PO4 , where ΔGh = Gibbs free energy change of
hydration) (9) and competitive hydrogen bond donation
associated with the solvent shell (10). Gabbaï and cowork-
ers have been successful in meeting this challenge. They
described the synthesis of a cationic Lewis acidic borane
which forms a highly stable zwitterionic ammonium/

SUPRAMOLECULAR CHEMISTRY
3
microorganisms (18) although the underlying microbiol-
ogy is still far from understood (19). Unfortunately, EBPR
is not likely to be applicable to medical applications, such
as treating hyperphosphatemia, where biocompatibility
is likely to be required.
Hyperphosphatemia, or elevated serum phosphate
levels, is a disorder that is particularly prevalent among
patients with end-stage renal disease (20). The current
methods for treating hyperphosphatemia involve various
combinations of long (3 h) daily dialysis session, dietary
restrictions, consumption of hard cations, or ingestion
of Sevelamer (a cross-linked poly(allylamine) (PAA), e.g.
Renagel®, c.f. Figure 2) (21).
Oral administration of cross-linked PAA serves to control
the concentration of phosphorus available for absorption
in the gastrointestinal track. This polymer is not selective for
phosphate. In fact, it also binds the chloride anion well. Rather
high doses must therefore be ingested. Adverse effects are
also seen in some patients undergoing Renagel® therapy. We
thus believe there remains a need for biologically compatible
materials capable of binding phosphate under conditions
associated with the treatment of hyperphosphatemia. More
broadly, there remains a need for polymers that can serve
as extractants for highly hydrated tetrahedral anions. Initial
efforts towards attaining these goals are summarised below.
We have previously reported a series of tetrakis(1H-
pyrrole-2-carbaldehyde) receptor family, including an
electroactive derivative and a hexakis(1H-pyrrole-2-carbal-
dehyde) derivative as excellent phosphate receptors (22,
23). Two of these macrocyclic Schiff-base structures, com-
pounds 1 and 2 (Figure 3), were chosen as models for the
development of polymeric materials containing pendent
2,6-diamidopyridine dipyrromethane (DPM) receptors as
detailed further below.
Figure 2. chemical structure of renagel®.
Polymeric materials with pendent phosphate recep-
tors were proposed based on a prior study wherein poly-
mers with pendent calix[4]pyrrole (C[4]P) anion receptors
(c.f. Figure 4) were used to extract halide anion salts from
aqueous phases. In that study, Aydogan and Sessler used a
meso-functionalized C[4]P system as a pendent receptor on
a poly-(methyl methacrylate) (pMMA) backbone to obtain
the co-polymer 3 (13). This co-polymer was shown to be
Figure 3. Structures of compounds 1 and 2.
Figure 4. Structures of key compounds from reference (14).

4
M. K. DELIOMEROGLU ET AL.
capable of extracting tetrabutylammonium (TBA) chloride
(TBACl) and fluoride (TBAF) salts from aqueous media into
an CH2Cl2 organic phase, under conditions where the con-
trol compounds 7 (octamethyl C[4]P) and 6 (a pMMA pol-
ymer without the pendant receptors) were not effective.
Aydogan and Sessler improved their first generation
C[4]P-pMMA material (3) by creating a co-polymer (4) that
also contained a pendent cation-binding receptor (benzo-
15-crown-5, 8) (14). Co-polymer 4, which contains both
pendent anion and cation receptors, was found capable of
extracting halide salts from water as ion-pairs, namely KF
and KCl, into an organic phase. The anion extraction could
be followed visually by using a cationic dye with a Cl− coun-
ter anion (9). This colored reagent was carried from the
aqueous solution into the organic phase (CH2Cl2) by the
polymer 4, as inferred from the observed color changes.
This was not true for any control systems, including mon-
omer 7, crown ether 8, the monofunctionalized polymer
3, pMMA 6 or combinations thereof (14). The extraction
events were also monitored by various analytical meth-
ods, e.g. 1H and 19F NMR spectroscopy and flame emission
spectroscopy. Against this background an effort was made
to functionalize the known phosphate anion receptors 1
and 2 such that they could be incorporated into polymers.
164.0, 159.1, 151.6, 150.4, 146.2, 143.7, 139.4, 138.7, 131.3,
131.0, 128.6, 127.7, 127.2, 126.4, 126.0, 119.6, 119.4, 114.3,
109.7, 55.7, 45.4, 29.5 ppm. HRMS (ESI+) m/z for C38H32N7O3
[M]+ calcd 634.25565, found 634.25611.
(4E,13E)-2-(4-Hydroxyphenyl)-2-methyl-11H,31H-
5,7,11,13-tetraaza-9(2,6)-pyridina-1,3(2,5)-dipyrrola-
6,12(1,2)-dibenzenacyclotetradecaphane-4,13-dien-
e-8,10-dione (22)
The procedure reported for the formation of the other
Schiff-base macrocycle 21 was followed, albeit with 15
and 20 as the starting materials. In a 100 ml round bot-
tom flask, compound 15 (352 mg, 1.1 mmol) and com-
pound 20 (347 mg, 0.98 mmol) were dissolved in distilled
methanol (60 ml). H2SO4 conc. (115 μL, 2.1 mmol) was then
added dropwise using a syringe. The resulting solution was
heated at reflux under argon for 15 min. The solution was
cooled to room temperature and an excess of TEA was
added via syringe. The solution was stirred at room tem-
perature for 15 min. The product was collected in the form
of a very pale yellow precipitate and washed with cold
methanol to yield 490 mg of macrocycle 22 (93% yield).
1H NMR (400 MHz, DMSO-d6) δ 10.99 (bs, 2H), 10.33 (s, 2H),
9.26 (s, 1H), 8.14–8.16 (m, 3H), 8.05 (s, 2H), 7.35 (d, 2H), 7.26
(t, 2H), 7.16 (t, 2H), 7.04 (d, 2H), 6.57–6.59 (m, 6H), 6.12 (d,
2H), 1.88 (s, 3H) ppm. 13C NMR (150 MHz, DMSO-d6) δ 163.2,
155.8, 151.5, 149.2, 149.0, 142.8, 139.3, 137.7, 130.7, 129.4,
128.1, 127.6, 127.3, 124.9, 124.4, 119.8, 117.3, 114.7, 108.7,
44.2, 29.7 ppm. HRMS (ESI+) m/z for C37H29N7O3 [M + H]+
calcd 620.24063, found 620.24046.
Experimental
Synthesis
(4E,13E)-2-(4-Methoxyphenyl)-2-methyl-11H,31H-
5,7,11,13-tetraaza-9(2,6)-pyridina-1,3(2,5)-dipyrrola-
6,12(1,2)-dibenzenacyclotetradecaphane-4,13-diene-
8,10-dione (21)
Co-polymer 24
In a 50 ml round bottom flask, 14 (46.4 mg, 0.144 mmol,
1 equiv.) and 20 (50 mg, 0.144 mmol) were dissolved
in dry methanol (20 ml). Concentrated H2SO4 (16.9 μL,
0.317 mmol, 2.2 equiv.) was then added using a syringe.
The resulting solution was heated at reflux under argon
for 15 min. The volatiles were removed under reduced
pressure. The residue obtained in this way was washed
with CHCl3 to give a salt-like orange-brown solid. This
crystalline solid was redissolved in CH3OH. An excess of
triethylamine (TEA; 0.1 ml) was added to this solution. The
resulting mixture was heated at reflux for 15 min then was
cooled to room temperature. The product was collected
in the form of a light pink precipitate, which was washed
with cold methanol to yield the free-base macrocycle. 1H
NMR (500 MHz, CD2Cl2) δ 1H NMR (500 MHz, CD2Cl2) δ 9.52
(bs, 2H), 8.27 (d, 2H, J = 7.8 Hz), 8.23 (s, 2H), 8.02 (t, 1H,
J = 7.9 Hz), 7.59 (dd, 2H, J = 7.8 Hz, J = 1.7 Hz), 7.24 (dt, 2H
J = 7.4 Hz, J = 1.6 Hz), 7.20 (dt, 2H, J = 7.4 Hz, J = 1.6 Hz), 7.11
(dd, 2H, J = 7.8 Hz, J = 1.7 Hz), 6.82 (d, 2H, J = 8.8 Hz), 6.74 (d,
2H, J = 3.8 Hz), 6.65 (d, 2H, J = 8.8 Hz), 6.33 (d, 2H, J = 3.8 Hz),
3.67 (s, 3H), 2.01 (s, 3H) ppm. 13C NMR (126 MHz, CD2Cl2) δ
To a solution of 22 (340 mg, 0.55 mmol) in DMF (5 mL), were
added Cs2CO3 (185 mg, 0.56 mmol) and CsBr (2 mg). The
resultant mixture was stirred for 15 min at room temper-
ature. A solution of the chloromethyl functionalized pol-
ymer 23 (85 mg, containing c.a. 0.55 mmol Cl atom) was
then added. The mixture was heated to 80 °C for 12 h. The
reaction mixture was then allowed to cool to room tem-
perature. Before being sparged into CH3OH, the insoluble
precipitates were removed by filtration. The filtrate was pre-
cipitated in CH3OH (200 mL). The precipitate that resulted
was filtered and dried in vacuo. The solid obtained was dis-
solved in CH2Cl2 and insoluble inorganics and non-reacted
22 were removed by filtration. CH2Cl2 was removed under
reduced pressure. Compound 24 was solidified with CH3OH
and dried in vacuo to give a light yellow powder (210 mg).
Compound 27
A similar procedure used to obtain co-polymer 24 was
employed, albeit with compounds 22 and 30 as start-
ing materials. The product (27) proved stable enough to
1
characterise but not to store on the bench top. H NMR

SUPRAMOLECULAR CHEMISTRY
5
(400 MHz, CDCl3) δ 9.51 (bs, 2H), 8.23 (d, J = 7.8 Hz, 2H), 8.19
(s, 2H), 7.91 (t, J = 7.8 Hz, 1H), 7.59 (dt, J = 7.2 Hz, J = 2.0 Hz,
2H), 7.17 (m, 4H), 7.04 (dd, J = 7.2 Hz, J = 2.0 Hz, 2H), 6.80
(d, J = 8.9 Hz, 2H), 6.68 (d, J = 3.7 Hz, 2H), 6.63 (d, J = 8.8 Hz,
2H), 6.29 (d, J = 3.7 Hz, 2H), 4.08 (dd, J = 11.0 Hz, J = 3.0 Hz,
1H), 3.76 (dd, J = 11.0 Hz, J = 5.8 Hz, 1H), 3.22 (m, 1H),
2.81 (t, J = 4.9 Hz, 1H), 2.64 (dd, J = 4.9 Hz, J = 2.7 Hz, 1H),
2.0 (s, 3H) ppm. 13C NMR (101 MHz, CDCl3) δ 163.5, 157.2,
150.7, 149.7, 149.7, 142.7, 138.9, 138.6, 130.6, 128.1, 127.1,
126.5, 125.9, 125.6, 125.6, 118.8, 118.6, 114.4, 109.2, 68.7,
50.0, 36.5, 31.3, 29.3 ppm. HRMS (ESI+) calcd for C40H33N7O4
[M + H]+ 676.26670, found 676.26750.
subsided. The reaction vessel was then allowed to cool
down to room temperature. At this point, the reaction
mixture was split into two equal parts. To one portion,
the cross-linker (EPI, 30) (40 mg) was added. To the other
one, precursor 27 (49 mg) along with EPI (30) (40 mg) was
added. Both mixtures formed gels within an hour. The
dried gel slabs, corresponding to 32 and 33, were ground
into small pieces for use in solid-liquid extraction studies.
Results and discussion
Two new Schiff-base macrocycles, 21 and 22, were con-
structed by cyclisation of the dialdehydes 14 and 15 with
diamine 20 via diimine formation (Scheme 1). The dial-
dehydes 14 and 15 and the diamine 20 were prepared
according to literature procedures (24, 26). All structures
were characterised by 1H and 13C NMR spectroscopy
Randomly cross-linked PAA·HCl (32) and the randomly
cross-linked PAA·HCl derivative with Schiff-base
pendants (33)
A 20% w/v solution of PAA·HCl (160 mg, 0.8 ml) was mixed
with a 10 M solution of NaOH (68 mg) until the exotherm
1
and high-resolution mass spectrometry (HRMS). H NMR
Scheme 1. Synthesis of Schiff-base macrocycles 21 and 22.

6
M. K. DELIOMEROGLU ET AL.
spectroscopy was used to monitor the formation of imine
bonds in the final step. Imine bond formation was con-
firmed by the absence of a signal at 9.4 ppm (the reso-
nance frequency for the aldehyde CH proton of 15) and
the presence of a signal at 8.3 ppm ascribed to the imine
CH proton (cf. Figure 5). Figure 6 shows the X-ray crystal
structures of two precursors 15 and 20 and the two Schiff-
base macrocycles 21 and 22.
While macrocycle 21 was designed as a control com-
pound, macrocycle 22 was designed to serve as a precursor
that would allow attachment of the Schiff-base macrocy-
clic framework to a polymeric support. In other words, it
was expected that macrocycle 22 would allow access to
polymeric materials with pendent Schiff-base receptors.
Solid state structural studies were carried out using
single crystals of 21 co-crystallized with CH3OH and 22
co-crystallized with DMF. Both structures revealed similar
conformations in which the aryl moiety of the DPM subunit
(originating from 15) and the pyridine moiety (originating
from 20) are in cis positions relative to the o-phenylene
rings. Based on the conformational similarity between
the single crystal structures of 21 and 22, it was inferred
that the methyl group did not have a substantial impact
on the conformation of the macrocyclic framework. This
lack of difference was considered to augur well for the use
of macrocycles functionalized in the phenolic position as
precursors for polymer formation.
It is also instructive to compare the conformations of
these two new Schiff-base macrocycles (21 and 22) with
those of similar macrocycles reported previously, namely
1 (24) and 2 (26). Reported modeling studies showed that
the free ligands 1 and 2 could exist in two different con-
formations (26). In the case of 2, the tolyl and pyridine
fragments were calculated to adopt either in trans or cis
positions relative to the o-phenylene rings. However, sin-
gle crystal structures of both 21 and 22 revealed a cis-like
conformation. The contrast was thought to arise from two
structural differences in the DPM subunits. Compound 2
has a tolyl ring, a hydrogen atom, and two fully β-substi-
tuted pyrrole rings in the meso-position, whereas the new
system 22 has an aryl ring, a methyl group, and two β-free
pyrrole rings.
It is likely that these structural differences, particularly
those involving the meso-positions and the β-substituents
of the pyrrole rings, are reflected in the actual conforma-
tions observed for macrocycles 1, 2, 21, and 22.
Figure 5. (colour online) Stacked partial 1h Nmr spectra of 15 (bottom), 20 (middle), and 22 (top) in cD3cN.

SUPRAMOLECULAR CHEMISTRY
7
Figure 6. (colour online) Single crystal structures of (a) 15, (b) 20, (c) 21·2ch3oh, and (d) 22·2DmF.
obtained from the titration studies, as well as the affinities
previously reported for 1 (24) and 2 (26) in CH3CN.
From the binding affinity constants listed in Table 1, the
Anion binding studies of 21 and 22
It was considered possible that the anion binding affinities
of the new receptor systems could differ from one another
and related prior systems because of these inferred struc-
tural differences. The anion binding affinities of the new
systems, 21 and 22, were thus tested.
−
order of anion selectivity of 21 for the oxyanions H2PO4
and HSO4− in CH2Cl2 is the same as that of the receptor 1
in CH3CN. However, the H2PO4− binding stoichiometries of
the two systems (21 and 1) differ. A 1:2 binding stoichiom-
Figure 7 shows partial 1H NMR spectra of 21 recorded
in CD2Cl2 in the presence of varying molar equivalents of
TBAH2PO4. Major changes in the chemical shifts of the
pyrrole NH signals and the imine CH signals are observed
(cf. Figure 7). Saturation behavior of the changes in the
shifts is observed after approximately 2.5 M equivalents
of TBAH2PO4 had been added to the initial receptor
solutions.
For quantification of the binding affinities, UV-vis
spectroscopic studies, in which receptor 21 was titrated
with TBAH2PO4 and TBAHSO4 in CH2Cl2, were conducted.
Experimental details and spectra are presented in the
Supplemental Material. Table 1 shows the binding affinities
−
etry for H2PO4− was reported for 1. Formation of 1·2H2PO4
complex was supported by a competition study in which
a solution of 1 containing 10 M equivalents of HSO4− was
titrated with H2PO4− (24). The spectral changes observed
in the case of 1 were interpreted in terms of the receptor
having a second binding site for H2PO4− that was effective
for dihydrogen phosphate anion complexation even in the
presence of 10 M equivalents of HSO4− (24).
Receptor 21, on the other hand, displayed a 1:1 binding
−
stoichiometry for H2PO4 anion in CH2Cl2. The isosbestic
point observed1 at 327 nm in the UV-vis spectral titration is
most easily interpreted in terms of two absorbing species at

8
M. K. DELIOMEROGLU ET AL.
Figure 7. (colour online) Stacked partial 1h Nmr spectra of 21 in cD2cl2 at varying molar ratio of tBah2po4.
Table 1. Binding affinity constants (in m−1) for 21, 22, 1, and 2 with tBa+ salts of oxyanions as determined by uV-vis spectroscopy in the
specified solvents.
Anion
21a
22b
1b,c
2b,d
−
h2po4
(2.1 0.3) × 105
1.7 × 105
K1 = 3.4 × 105
(2.9 0.2) × 104
K2 = 2.6 × 104
−
hSo4
(1.2 0.2) × 104
nd
(6.4 0.3) × 104
(1.1 0.2) × 105
ain ch2cl2.
bin ch3cN.
creported in reference-(24).
din reported in reference-(26), nd: not determined.
this wavelength (namely, the free receptor 21 and the bound
form of the receptor, 21·H2PO4 ) that are interconverting
One of the most substantial differences between the
two new receptors, 21 and 22, was their relative solubili-
ties. In fact, derivative 22, the putative precursor for poly-
meric materials, displayed low solubility in organic solvents
(e.g. CH2Cl2, CHCl3, CH3OH, and CH3CN), whereas 21 was
freely soluble at the mM 006Cevel in typical halogenated
solvents. The low solubility of 22 precluded quantitative
analyses in CH2Cl2. It could be studied in CH3CN, the same
solvent in which 1 and 2 were studied. The affinity con-
stant of 22 for H2PO4− was determined (Ka = 1.7 × 105 M−1
via UV-vis spectroscopy and Ka = 1.6 × 105 M−1 via ITC)
and proved competitive with those of previous recep-
tors. Although it was intended that structure 22 would
be further modified at the phenolic position to allow its
development as a polymer precursor, it was gratifying that
the core macrocycle proved effective for H2PO4− binding.
−
between one another. A Job plot obtained from the same
experiment likewise proved a 1:1 binding stoichiometry.
Additionally, the data obtained from UV-vis spectra could
be fit well to a 1:1 binding isotherm. Based on these obser-
vations, we conclude that receptor 21 has high affinity and
selectivity for the H2PO4− anion and binds this target with
a 1:1 binding stoichiometry in organic media (i.e. CH2Cl2).
Among the receptors listed inTable 1, macrocycle 2 dis-
−
plays the greatest affinity for HSO4 . This was rationalised
in terms of its ‘frozen’ conformation, which presumably
results from the fact that the β-positions of the pyrrole
units are substituted and that a sterically restricting tolyl
group is present on one of the ‘meso-like’ positions of the
DPM subunit.

SUPRAMOLECULAR CHEMISTRY
9
(or to a polymerizable moiety) before formation of the
Schiff-base macrocycle (Scheme 2). The second approach
involved attachment of pre-formed Schiff-base systems
(e.g. 22) to a polymer backbone (or to a polymerizable
moiety) (Scheme 3).
Synthesis of polymeric material with pendent
schiff-base macrocycles
In order to obtain polymeric materials containing pen-
dent Schiff-base macrocycles, two general synthetic
approaches were considered. The first approach involved
attachment of DPM subunits to a polymer backbone
The approach involving the attachment of a preformed
functionalized Schiff-base to a functionalized polymer
backbone was tested first. The functionalized Schiff-base
22, prepared in 93% yield (see Experimental Section), was
attached to the commercially available functionalized-pol-
ystyrene 23 (poly(vinylbenzyl chloride), 60:40 mixture
of the 3- and 4- isomers) under SN2 reaction conditions
(Scheme 4).
1
As inferred from integration of the H NMR spectrum
(the ratio of the signal at 9.5 ppm to those at 9–5.5 ppm),
the new polymeric material 24 bears one pendent Schiff-
base macrocycle for approximately every 3.5 phenyl
groups present on the styrene backbone (cf. Figure 8). In
other words, the reaction between equimolar amounts of
22 and 23 yielded about 30% substitution of 23 with 22.
The net result is a co-polymer, 24, bearing pendent Schiff-
base macrocycles (Scheme 4). Co-polymer 24 proved
appreciably soluble in CHCl3. It could thus be used for liq-
uid-liquid and solid-liquid extraction studies as detailed
below.
Scheme 2. Firstsyntheticapproachconsidered forthesynthesis of
polymeric materials containing Schiff-base macrocyclic pendants.
it starts from diupyrromethane precursors.
Anion binding studies of co-polymer 24 in CHCl3 via
UV-vis spectroscopy
UV-vis spectroscopy was used to confirm the presence of
the pendent Schiff-base macrocycles in the case of co-pol-
ymer 24. Polymer 23, lacking absorbance features in the
300–400 nm spectral range, reacted with compound 22,
Scheme 3. Second synthetic approach considered for the
synthesis of polymeric materials containing Schiff-base
macrocyclic pendants. this approach relies on the use of
precursors containing the preformed macrocycle.
Scheme 4. preparation of polymer 24 by means of post-polymerisation modification of the commercially available functionalized
polystyrene 23 with 22.

10
M. K. DELIOMEROGLU ET AL.
1
Figure 8. (colour online) h Nmr spectra of 21 in cD2cl2 (bottom), commercially available functionalized polystyrene 23 in cDcl3
(middle), and co-polymer product 24 in cD2cl2 (top).
which is insoluble in CHCl3. The product, assumed to be
co-polymer 24, proved soluble in CHCl3. Moreover, it dis-
played UV-vis spectral features between 300 and 400 nm
in CHCl3 that are characteristic of a Schiff-base macrocycle
such as 22 (cf. Figure 9(a)).
involved a liquid-liquid extraction. In this case, a solution
of co-polymer 24 in CDCl3 was layered with a solution of
TBAH2PO4 in D2O (Figure 10). After shaking2 and separating
−
the phases, changes in the H2PO4 anion concentration
in both phases were quantified via 31P NMR spectroscopy
using either phenylphosphonic acid (PhPO3H) (in D2O) or
TBAPF6 (in CDCl3) as the internal standards.
Solutions of 24 in CHCl3 were screened for spectral
changes when treated with various TBA+ anion salts.
Among the halide anions, F− induced the biggest spec-
tral change (Figure 9(b)). The spectrum of 24 that resulted
upon exposure toTBAH2PO4 was similar to those produced
when 21 was treated with H2PO4− (cf. Figure 9c and Figure
S.5). On the other hand, the addition of TBAHSO4 addition
caused a spectral change in 24 that differed from what was
observed when 21 was exposed to the same salt (Figure
9(c)). The effect of TFA on the spectrum of 24 in CHCl3 was
also tested. The spectral change (Figure 9(d)) was attrib-
uted to protonation of the two imine groups present in
the Schiff-base macrocycle framework.
The second method involved solid-liquid extraction
studies. Here, solid co-polymer 24 was added into a
solution of TBAH2PO4 in CD3OD (or in D2O in a separate
experiment). After shaking and separating off the solution
−
phase, the change in the H2PO4 anion concentration in
CD3OD was quantified via 31P NMR spectroscopy using
PhPO3H as the internal standard. The results are summa-
rised in Table 2. These results proved hard to interpret. The
common result among four entries was an increase in the
−
concentration of H2PO4 in the source phase (i.e. stock
H2PO4− solution that was tested for extraction). Although
removal of H2PO4− from its original source phase was the
−
desired outcome, the unexpected increase in H2PO4 con-
Extraction studies with co-polymer 24
centration in the source phase can be explained by the
−
−
Copolymer 24 was tested for its ability to extract H2PO4
removal of water from the H2PO4 source phase, leading
anion salts using two experimental techniques. The first
to a more concentrated H2PO4− solution.

SUPRAMOLECULAR CHEMISTRY
11
Figure 9. (colour online) uV-vis spectra in chcl3 of (a) the polymer 23 (7.5 mg/l) and co-polymer 24 (18.6 mg/l), (b) 24 in the presence
of various tBa+ halide salts, (c) 24 in the presence of various tBa+ anion salts, (d) addition of tFa into a solution of 24.
Table 2. results of extraction studies conducted with co-polymer
24 as a putative extractant.
Extractant
solvent
−
Source solvent
% Change ([H2PO4 ])
Extraction
type
−
[H2PO4 ]a, V
[24], V
cDcl3
In source In extractant
liquid-liquid
D2o
50 mm, 1.0 ml 10 mg/1.0 ml
+13%
0%
Figure 10. (colour online) Graphical presentation of the liquid-
liquid extraction approach used to test co-polymer 24 as a
possible phosphate anion extractant (using tBa+ as the anion
source). left: h2po4 solution in D2o, right: Biphasic mixture
obtained by adding a solution of co-polymer 24 in cDcl3 to the
solution on the left.
Solid-liquid
cD3oD
20 mm, 1.5 ml
50 mm, 1.5 ml
D2o
Solid 24
10 mg
−6%
+4%
–
–
10 mg
Solid 24
11 mg
−
Solid-liquid
50 mm, 1.0 ml
+6%
–
−
atBah2po4 was used as the h2po4 source. V stands for volume. % changes
were quantified by using internal standard (phpo3h or tBapF6) via 31p Nmr
spectroscopy.
Figure 11 shows the UV-vis spectra of 24 in CHCl3 before
exposing to water, after shaking with water, and after
shaking with an aqueous TBAH2PO4 solution. The major
spectral difference was the change observed in the relative
intensity of the two maxima of the co-polymer 24 after
contacting with water. This was seen independent of the
H2PO4− content in the source phase.

12
M. K. DELIOMEROGLU ET AL.
During the liquid-liquid extraction study conducted
its phenolic position with a polymerizable styrene unit, a
clickable propargyl group, and a cross-linker, epichloro-
hydrin (EPI). This gave compounds 25, 26, and 27, respec-
tively (Scheme 5).
with 24, gelation behavior was seen regardless of the con-
tent of the aqueous solutions. On the basis of this observed
gelation behavior and the UV-vis spectral changes seen
when 24 (in CHCl3) is brought into contact with an aque-
ous phase, we propose that this polymeric material can
absorb water. Removing water from the aqueous phase
under conditions of liquid-liquid would serve to increase
the effective TBAH2PO4 concentration in the source phase,
thus accounting for the results in Table 2.
In order to incorporate the new macrocyclic frame-
work into a cross-linked PAA network, the synthetic route
shown in Scheme 6 was followed. The control system (32),
an EPI (30) cross-linked PAA, was prepared along with a
pendent Schiff-base containing derivative, 33. Both hydro-
gels were prepared via an aqueous reaction of a 20% w/v
solution of linear PAA·HCl chains and EPI, which served
as the crosslinking agent (27). The commercially availa-
ble PAA·HCl (Alfa Aesar) used in the synthesis of the gels
had molecular weights of 120,000–200000 g/mol. Before
crosslinking, the polymer chains, some of the HCl groups
of PAA·HCl were neutralised by adding NaOH solution. This
provides free NH2 sites for the subsequent EPI cross-link-
ing reaction. For the synthesis of 33, Schiff-base 27 was
added along with EPI in the cross-linking step. Two insol-
uble systems, 32 and 33, resulted. Both were tested as
possible H2PO4− extractants using a solid-liquid extraction
approach.
Synthesis of cross-linked polymeric material with
pendent schiff-base macrocycles
Since the attachment of 22 to 23 to obtain co-polymer 24
using substitution chemistry (vide supra) worked well, vari-
ous other modifications in the phenolic position of 22 were
tested. Macrocycle 22 was substituted regioselectively in
Extraction studies with 33
Figure 13 shows three 31P NMR spectra arising from the
solid-liquid extraction studies in which 32 and 33 were
used in conjunction with KH2PO4. The bottom spectrum
is of a stock KH2PO4 solution (46 mM) in D2O containing
phenylphosphonic acid (PhPO3H, 33 mM) as the internal
standard.
The spectrum in the middle is of the KH2PO4 (1.0 mL)
aliquot after it had been treated with control cross-linked
−
PAA 32 (15 mg). A decrease in H2PO4 concentration
from 46 to 35 mM was observed. The top spectrum is
of a separate KH2PO4 (1.0 mL) aliquot after having been
treated with Schiff-base appended cross-linked PAA 33
(15 mg). A decrease in H2PO4− concentration from 46 mM
Figure 11. (colour online) uV-vis spectra of 24 (25 mg/l) in chcl3
(black), 24 in chcl3 after vigorous shaking with deionized (Di)
water (red), and 24 in chcl3 after shaking with aqueous tBah2po4
solution (green).
Figure 12. (colour online) photos of gel formed during the extraction study with the co-polymer 24. the bottom layer initially consisted
of a solution of the co-polymer 24 in chcl3 while the top layer was an aqueous anionic salt solution.

SUPRAMOLECULAR CHEMISTRY
13
Scheme 5. Syntheses of 25–27 via nucleophilic substitution.
Scheme 6. Synthetic route used to prepare a cross-linked polymeric material with pendent Schiff-base macrocycles (33). also shown is a
synthesis of a control system without the pendant macrocycles (32).
to 38 mM was observed. On this basis, it was concluded
benefit associated with attaching a receptor to the pol-
ymer system. While disappointing, this result serves to
underscore the need for more effective receptor design
in the construction of polymeric materials designed
to promote the extraction of anionic salts into organic
media.
that the functionalized solid hydrogel 33 (15 mg) was
−
no more effective at extracting H2PO4 than the same
amount of solid hydrogel 32, after accounting for differ-
ences in effective molecular weights. On this basis, we
conclude that electrostatic effects override any putative

14
M. K. DELIOMEROGLU ET AL.
Figure 13. (colour online) Stacked partial 31p Nmr spectra (recorded in D2o) of a stock Kh2po4 solution (46 mm) containing
phenylphosphonic acid (phpo3h, 33 mm) as the internal standard (bottom), the stock aqueous Kh2po4 solution (1.0 ml) after treatment
with cross-linked paa 32 (15 mg) containing phpo3h (33 mm) as the internal standard (middle), and the stock aqueous Kh2po4 solution
(1.0 ml) after treatment with the Schiff-base appended cross-linked paa 33 (15 mg) containing phpo3h (33 mm) as the internal standard
(top).
in the source aqueous solution when 24 was tested as an
extractant.
Conclusion
A new functionalized Schiff-base macrocyclic framework,
which relies on a formylated-DPM as a precursor, was pre-
pared. Control compound 21, bearing this new framework,
was studied for its anion binding properties in CH2Cl2. It was
concluded that 21 displays higher affinity for H2PO4− than
it does for HSO4− in CH2Cl2. In this solvent, the H2PO4− anion
A hydrogel containing pendent Schiff-base macro-
cycles (33) was also prepared. In hydrogel 33, pendent
Schiff-bases were covalently attached to a PAA backbone
by adding a mixture of 27 and EPI during the cross-link-
ing step. Solid-liquid extraction studies revealed that 33
is able to extract KH2PO4 into an organic phase, but is not
efficient in doing so than a control hydrogel 32, a result
that is rationalised in terms of non-specific electrostatic
interactions dominating over receptor-based binding in
the case of the present systems. Current work is focused
on preparing polymers bearing functionality designed to
enhance anion and ion pair recognition.
−
is bound with a binding stoichiometry 1:1 (21: H2PO4 ).
Covalent attachment of the new framework to a polymeric
backbone was achieved using derivative 22. It was shown
that 22 could be attached to many different systems via
substitutions involving the phenolic position of 22. In this
way, derivatives 25–27 were prepared. The co-polymer 24,
also obtained via a substitution reaction, was studied for its
anion binding potential in CHCl3 via UV-vis spectroscopy
and for its extraction ability. Little change in the feature of
co-polymer 24 was seen in CHCl3 with all anions tested,
Notes
1. The total concentration of the receptor species was held
constant throughout the titration of 21 with TBAH2PO4.
2. Shaking a biphasic mixture co-polymer 24 in CDCl3 and
TBAH2PO4 in D2O causes gelation. Gelation occurs in the
bottom layer (i.e. CDCl3). A photo of the resulting gel is
shown in Figure 12.
including the common halides, H2PO4 , and HSO4− (all as
−
their TBA+ salts). Under conditions of liquid-liquid extrac-
tion, co-polymer 24 forms gels, independent of whether
an anion is present in the aqueous phase. Extraction stud-
ies did not reveal a reduction in the H2PO4− concentration

SUPRAMOLECULAR CHEMISTRY
15
(12) Hirai, M.; Gabbaï, F.P. Angew. Chem. Int. Ed. 2015, 54, 1205–
1209.
(13) Aydogan, A.; Coady, D.J.; Lynch, V.M.; Akar, A.; Marquez, M.;
Bielawski, C.W.; Sessler, J.L. Chem. Commun. 2008, 1455–
1457.
(14) Aydogan, A.; Coady, D.J.; Kim, S.K.; Akar, A.; Bielawski, C.W.;
Marquez, M.; Sessler, J.L. Angew. Chem. Int. Ed. 2008, 47,
9648–9652.
(15) Wang, Z.; Luecke, H.; Yao, N.; Quiocho, F.A. Nat. Struct. Biol.
1997, 4, 519–522.
Disclosure statement
No potential conflict of interest was reported by the authors.
Funding
This work was supported by the U.S. Department of Ener-
gy, Office of Basic Energy Sciences [grant number DE-FG02-
01ER15186 to J.L.S.].
(16) Levin, G.V.; Shapiro, J. Water Environ. Fed. 1965, 37, 800–
821.
(17) Barnard, J.L. Water Res. 1975, 9, 485–490.
(18) Mino, T.; van Loosdrecht, M.C.M.; Heijnen, J.J. Water Res.
1998, 32, 3193–3207.
ORCID
Jonathan L. Sessler
http://orcid.org/0000-0002-9576-1325
(19) Mino, T. Biochemistry. (Mosc). 2000, 65, 341–348.
(20) Goldsmith, D.R.; Scott, L.J.; Cvetković, R.S.; Plosker, G.L.; Bihl,
G.; Kidney, W.; West, S.; Africa, S.; Bommer, J. Drugs 2008,
68, 85–104.
(21) Ayus, J.C.; Achinger, S.G.; Mizani, M.R.; Chertow, G.M.;
Furmaga, W.; Lee, S.; Rodriguez, F. Kidney Int. 2007, 71,
336–342.
(22) Deliomeroglu, M.; Lynch, V.; Sessler, J. Chem. Commun.
2014, 50, 11863–11866.
(23) Deliomeroglu, M.; Lynch, V.; Sessler, J. Chem. Sci. 2016, 7,
3843–3850.
(24) Sessler, J.L.; Katayev, E.; Pantos, G.D.; Ustynyuk, Y.A. Chem.
Commun. 2004, 1276–1277.
(25) Lee, C.H.; Na, H.K.; Yoon, D.W.; Won, D.H.; Cho, W.S.; Lynch,
V.M.; Shevchuk, S.V.; Sessler, J.L. J. Am. Chem. Soc. 2003,
125, 7301–7306.
(26) Sessler, J.L.; Katayev, E.; Pantos, G.; Scherbakov, P.;
Reshetova, M.D.; Khrustalev, V.N.; Lynch, V.M.; Ustynyuk,
Y.A. J. Am. Chem. Soc. 2005, 127, 11442–11446.
(27) Kioussis, D.R.; Smith, D.F.; Kofinas, P. J. Appl. Polym. Sci.
2001, 80, 2073–2083.
(28) Lee, C.-H.; Lindsey, J.S. Tetrahedron 1994, 50, 11427–11440.
(29) Varghese, S.; Senanayake, T.; Murray-Stewart, T.; Doering,
K.; Fraser, A.; Casero, R.A.; Woster, P.M. J. Med. Chem. 2008,
51, 2447–2456.
References
(1) Sessler, J.L.; Gale, P.A.; Cho, W.-S. Anion Receptor Chemistry;
The Royal Society of Chemistry: Cambridge, UK, 2006.
(2) Steed, J.W.; Turner, D.R.; Wallace, K. Core Concepts in
Supramolecular Chemistry and Nanochemistry; Wiley,
Hoboken, NJ, 2007.
(3) Rashin, A.; Honig, B. J. Phys. Chem. 1985, 89, 5588–5593.
(4) Sessler, J.L.; Cyr, M.J.; Lynch, V.; McGhee, E.; Ibers, J.A. J. Am.
Chem. Soc. 1990, 112, 2810–2813.
(5) Katayev, E.A.; Boev, N.V.; Khrustalev, V.N.; Ustynyuk, Y.A.;
Tananaev, I.G.; Sessler, J.L. J. Org. Chem. 2007, 72, 2886–
2896.
(6) Seidel, D.; Lynch, V.; Sessler, J.L. Angew. Chem. Int. Ed. 2002,
41, 1422–1425.
(7) Jeffrey, G.A. An Introduction to Hydrogen Bonding (Topics in
Physical Chemistry); Oxford University Press, 1997.
(8) Camiolo, S.; Gale, P.A.; Hursthouse, M.B.; Light, M.E.; Shi, A.J.
Chem. Commun. 2002, 3, 758–759.
(9) Marcus, Y.J. J. Chem. Soc., Faraday Trans 1991, 87, 2995–
2999.
(10) Schmidtchen, F.P.; Berger, M. Chem. Rev. 1997, 97, 1609–
1646.
(11) Chiu, C.W.; Gabbaï, F.P. J. Am. Chem. Soc. 2006, 128, 14248–
14249.